Every Smelteon engagement produces the same artefact: a defensible before-and-after on first-pass yield, scrap, porosity and melt energy, measured against a baseline captured in shadow mode.
[ASPIRATIONAL / PLACEHOLDER] Smelteon is pre-revenue. The engagements below are modelled design-partner scenarios that describe the measurement method we commit to, not audited customer results. Real, named case studies will replace them once design partners approve publication.
Built with foundries and die casters melting iron, steel, aluminium, magnesium and copper
[PLACEHOLDER] Design-partner names are illustrative until first references are signed.
A 180kt/yr ductile iron foundry running an automotive crankshaft programme. Pour consistency varied by shift, misrun and cold-shut scrap concentrated on one geometry, and two of three experienced pourers were within a year of retirement.
A structural aluminium caster ramping an EV programme where CT inspection could not keep pace with output. Sampling left escape risk, and every reject destroyed hours of accumulated value.
A multi-alloy foundry where scrap-charge variation drove chemistry drift and correction additions, melt energy dominated the plant's utility bill, and inoculation fade quietly changed metallurgy between pours.
Every engagement uses the same measurement discipline, agreed in writing before the pilot starts.
| Stage | What happens | Who signs |
|---|---|---|
| Baseline | Shadow-mode instrumentation over a representative production period using the plant's own records as ground truth | Plant director, metallurgist |
| Golden set | Agent predictions scored against the plant's inspection results for that alloy and geometry family | Metallurgist, quality engineer |
| Assist | Recommendations approved by an engineer; outcomes graded by inspection; corrections captured as labels | Metallurgist |
| Report | Baseline vs. improved KPI report on yield, scrap, porosity, melt energy and takt | Plant director, finance |
| Safety review | Process windows, rollback paths, escalation rules and robot envelopes reviewed | Quality, EHS |
| Bounded | Low-risk cells graduate to adaptive control with exceptions escalating | Plant director, EHS |
A defensible before-and-after on first-pass yield, scrap rate, porosity rate, melt energy and takt at the wedge cell.
A documented, working integration with your furnace, molding, pouring, inspection and MES systems — including failure modes.
A labelled golden set for your alloy and geometry family, built from your own inspection ground truth.
The immutable action, approval and outcome log, tied to casting genealogy, exportable for IATF 16949 or AS9100 review.
8 wks
Typical shadow-mode baseline period
3
Autonomy gates before write-back
1
Wedge cell per engagement
0
Results published without customer approval
As design partners graduate to assist and bounded modes, these are the studies we intend to publish — with their names on them, and their numbers.
In progress · design partner
Misrun and cold-shut scrap, pour consistency across shifts and takt recovery on a moving green-sand line.
In progress · design partner
Detection accuracy against plant CT verdicts and value recovered before secondary machining operations.
Scoping
Chemistry-in-window rate, correction additions and kWh per tonne across scrap-charge variation.
Scoping
Finishing consistency, rework rate and hours of hot, dusty manual exposure removed from the plant.
“We will publish a customer's number when the customer signs off on it, and not one quarter before.”
Design partners get preferential expansion terms in exchange for case-study rights once the numbers are in.